Webinar: Improving Walkability at Signalized Intersections ...Webinar: Improving Walkability at...
Transcript of Webinar: Improving Walkability at Signalized Intersections ...Webinar: Improving Walkability at...
Portland State UniversityPDXScholar
TREC Webinar Series Transportation Research and Education Center(TREC)
1-26-2017
Webinar: Improving Walkability at Signalized Intersections withSignal Control StrategiesEdward J. SmaglikNorthern Arizona University
Sirisha Murthy KothuriPortland State University
Let us know how access to this document benefits you.Follow this and additional works at: http://pdxscholar.library.pdx.edu/trec_webinar
Part of the Transportation Commons, Transportation Engineering Commons, and the UrbanStudies Commons
This Book is brought to you for free and open access. It has been accepted for inclusion in TREC Webinar Series by an authorized administrator ofPDXScholar. For more information, please contact [email protected].
Recommended CitationSmaglik, Edward J. and Kothuri, Sirisha Murthy, "Webinar: Improving Walkability at Signalized Intersections with Signal ControlStrategies" (2017). TREC Webinar Series. 16.http://pdxscholar.library.pdx.edu/trec_webinar/16
Improving Walkability at Signalized Intersections with Signal Control StrategiesContributors: Sirisha Kothuri, Edward Smaglik, Andrew Kading, Chris Sobie, Peter Koonce, Paul Zebell, Avery Rhodes, Jeffrey Bauman Organizations:Portland State University, Northern Arizona University, City of Portland, OR, City of Mesa, AZ and City of Flagstaff, AZNITC Webinar, January 26th, 2017
Outline• Introduction• Background• Goals• Pedestrian Control Strategies• Pedestrian Priority Algorithm Development & Deployment• Conclusions and Recommendations
Agenda – Findings – Discussion – Next Steps
Introduction
Source: Bicycling and Walking in the United States; 2014 Benchmarking Report
Introduction Delays affect pedestrians
disproportionately “Everyone is a pedestrian”
Increa
singP
riority
How do we translate “pedestrian first” policies into specific operational strategies at intersections?
Background• Limited knowledge regarding signal control strategies focused on pedestrians
– Existing strategies typically focus on safety• Leading Pedestrian Interval (LPI)• Exclusive Pedestrian Phase (Barnes Dance)
– Other efficiency-focused options• Shorter cycle length• Actuated-coordinated timing• Free operation
Project Goals• Assess the efficiency impacts of existing strategies
– Coordination (base case)– Leading Pedestrian Interval– Exclusive Pedestrian Phase – Shorter Cycle Lengths– Free Operation– Actuated-Coordinated Operation
• Develop and implement a pedestrian priority algorithm
Leading Pedestrian Interval
Leading Pedestrian Interval - Impacts• Safety Impacts
– Documented reduction in pedestrian-vehicle crashes at intersections– Can provide level of ‘comfort’ to pedestrians
• Efficiency Impacts– Increase in overall delays due to lost time– Actual magnitude of increase depends on
• Length of LPI• Cycle length• Implementation of LPI on major or minor phase
Leading Pedestrian Interval -Implementation• Implementation should be based on (Sainenejad and Lo, 2015, Sharma et al., 2017)
– Crash history (frequency, severity)– Volume of pedestrians– School proximity– Activity by elderly residents– Impacts on vehicle delay– Visibility issues– Intersections with special geometry
Exclusive Pedestrian Phase
Exclusive Pedestrian Phase- Impacts• Safety Impacts
– Documented reduction in pedestrian-vehicle crashes at intersections– Increase in pedestrian signal non-compliance– Pedestrians less likely to wait for exclusive phase at low volume intersections
• Efficiency Impacts– Significant increase in overall delays due to lost time for all users– Actual magnitude of increase depends on
• Length of EPP• Cycle length
Exclusive Pedestrian Phase-Implementation• Best suited for
– Intersections with high volume of pedestrians and turning vehicles ( e.g. downtown)– Locations where traditional pedestrian accommodations do not work well– Carefully weigh the costs
• Increase in overall delays• Increase in pedestrian non-compliance
Free Operation• Each intersection operates independently of adjacent intersections• Individual intersections can be optimized without consideration of other signals• Can lead to greater flexibility and responsiveness (Urbanik et al., 2015)• Good detection is necessary on all approaches for high operational and safety performance (Koonce et al., 2008)
Free Operation - Impacts• Safety Impacts
– None quantified in literature• Efficiency Impacts
– Can reduce overall delays when major street vehicular volumes are low (v/c < 0.5)– Tradeoff between major and minor street user delays with higher volumes
• Reduction in minor street pedestrian delay• Increase in major street vehicle delay
– Detrimental to coordination
Free Operation - Implementation• Best suited for
– Locations with long spacing between adjacent intersections– Locations where coordination is not a priority– Time of day operation to prioritize pedestrians
• Off-peak (middle of day)• Late night
– Intersections with balanced volumes on major and minor streets
Short Cycle Lengths• Cycle length refers to the time taken for a complete sequence of signal indications• Ped delay is a function of cycle length• According to the HCM
ೢೌೖ మ
WhereC = cycle lengthgwalk = effective walk time
Short Cycle Lengths- Impacts• Safety Impacts
– None quantified in literature• Efficiency Impacts
– Reduce pedestrian delay– Encourages signal compliance– Increases efficiency of all users, typically
Short Cycle Lengths - Implementation• Best suited for off-peak and other low vehicular demand periods, when agencies want to keep their signals coordinated• NACTO’s Urban Street Design Guide and PEDSAFE also recommend short cycle lengths to increase compliance and efficiency
Actuated-Coordination• Actuating the coordinated phase allows
– Coordinated phases to gap out if there is low demand– Signal to be more responsive to field conditions
Source: Signal Timing Manual, 1st Edition
Actuated-Coordination - Impacts• Safety Impacts
– None quantified in literature• Efficiency Impacts
– Decrease in v/c ratios and fewer occurrences of split failures (Day et al., 2008)– Use of fixed force-offs and fully actuated coordination reduced delays for non-coordinated phases (Day et al., 2014)– Decreases minor street pedestrian delay (Sobie et al., 2016)
Actuated-Coordination - Implementation• Best suited for
– Off-peak and other low vehicular demand periods, when agencies want to keep their signals coordinated– Major street demand is low and minor street demand is high
• Presence of mainline detection is necessary– Additional detection and maintenance costs
Permissive Length• Period of time after the yield point where the call on a non-coordinated phase can be serviced without delaying the start of the coordinated phase• Increasing permissive length can reduce pedestrian and vehicular delays on the minor street movements
Permissive Length- Impacts• Safety Impacts
– None quantified in literature• Efficiency Impacts
– For low volume conditions, increasing permissive length reduces delays for non-coordinated phases (de Castro-Neto, 2005)
– Decreases minor street pedestrian delay (Kothuri et al., 2013)
Permissive Length - Implementation• Best suited for
– Off-peak and other low vehicular demand periods, when agencies want to keep their signals coordinated– Major street demand is low and minor street demand is high
Pedestrian Priority Algorithm• Goal is to change the operational pattern based on volume input
– Coordinated above a threshold– Pedestrian plan below threshold
• Select from previously presented options, or other• Toe the line between vehicular and pedestrian objectives
Pedestrian Priority Algorithm Development
Threshold Determination
Simulated Network
Simulated Network
Algorithm Simulation Results
Algorithm Field Deployments• Field Deployments
– Raspberry Pi• Mesa, AZ• Flagstaff, AZ
– ASC/3 Controller• Portland, OR
Threshold Determination
Coordinated Operation
Ped Priority Plan
Site 1 – Mesa, AZ (Raspberry Pi)
Site 2 – Flagstaff, AZ (Raspberry Pi)
Site 3 – Portland, OR (ASC/3)
Deployment Outcomes• Mesa
– Reductions in pedestrian delay seen during off-peak typically coordinated time periodsTwo weeks, before (weekdays)
Count of Cars Average Travel Time (s) Count of Cars Average Travel Time (s) Count of Pedestrians Call Average Pedestrian Delay0000-0630 (C=90) 145 52.87 162 31.70 13 00:54.10630-1500 (C=100) 1780 55.90 1098 58.95 330 00:49.11500-1830 (C=110) 809 60.12 959 62.63 200 00:50.91830-0000 (C=90) 917 57.50 766 64.00 153 00:44.4
Two weeks, after (weekdays)Count of Cars Average Travel Time (s) Count of Cars Average Travel Time (s) Count of Pedestrians Call Average Pedestrian Delay
0000-0630 (C=90) 103 43.50 80 47.53 7 00:13.00630-1500 (C=100) 1534 50.95 895 56.96 317 00:22.01500-1830 (C=110) 768 54.83 827 59.94 169 00:52.01830-0000 (C=90) 650 58.51 678 62.70 147 00:26.0
WB EBTravel Time Pedestrian Data
Travel Time Pedestrian DataWB EB
Deployment Outcomes• Flagstaff
– Results less clear– Reduction in ped delay during weekday off peak, but other numbers are contradictory (PM plan)– Detection may have been an issue
One week, before (Weekdays) Count of Pedestrians Call Average Pedestrian Delay0000-0715 (C=Free) 72 0:00:350715-0815 (C=90) 140 0:00:230815-1630 (C=95) 394 0:00:411630-1730 (C=75) 73 0:00:191730-0000 (C=Free) 143 0:00:26One week, after (Weekdays) Count of Pedestrians Call Average Pedestrian Delay0000-0715 (C=Free) 94 0:00:370715-0815 (C=90) 78 0:00:400815-1630 (C=95) 591 0:00:341630-1730 (C=75) 52 0:00:481730-0000 (C=Free) 206 0:00:36
Deployment Outcomes• Portland
– Minor reductions in ped delay observed– Further refinement of threshold, my yield further reduction in ped delay
Results Ped 4 Ped 8n Avg. Ped Delay n Avg. Ped Delay
Without Algorithm (All Day) 1604 0:00:51 1617 0:00:51With Algorithm (All Day) 1590 0:00:48 1602 0:00:49Without Algorithm (10:00 -13:00) 285 0:00:51 283 0:00:53With Algorithm (10:00 - 13:00) 229 0:00:50 273 0:00:49
Comparison of Strategies• Pedestrian Delay
– Barnes Dance– Coordination– Actuated Coordination– Leading Pedestrian Interval– Short Cycle Lengths– Free
• Vehicle Delay– Barnes Dance– Short Cycle Lengths– Leading Pedestrian Interval– Actuated Coordination– Coordination– Free
More Delay
Less Delay
Conclusions• LPI and Barnes Dance may improve safety, but increase delays overall• Free operation and shorter cycle lengths can reduce pedestrian delays, but impacts on safety are not known• Field deployments corroborate simulation results• Generally no right solution. Implementation choice based upon operational objectives and intersection characteristics
Recommendations• Common needs for deployment of specific strategies
– Barnes Dance: Very high pedestrian or right turning volumes creating need to completely separate modes– LPI/Split LPI: Geometry or vehicle volumes that cause issues for pedestrians entering crosswalk– Reduced Cycle Length: Satisfy need to operate in coordinated mode while trading green bandwidth for reduced delays for other users
Recommendations• Common needs for deployment of specific strategies
– Free: Most aggressive strategy resulting in lower delay for most users at expense at main street traffic– Actuated-Coordinated: Can provide earlier minor street Walk if possibility of main street vehicle gap out. Useful for low-moderate pedestrian volumes
Publications• NITC Final Report
– http://ppms.trec.pdx.edu/media/project_files/NITC-RR-782_Final_Report.pdf• Transportation Research Record
– Sobie, Chris, Smaglik, Edward J., Sharma A., Kading A., Kothuri, S., and Koonce, P. “Managing User Delay with a Focus on Pedestrian Operations” Accepted for publication in Transportation Research Record, in press, 2016.– Conference Paper: http://docs.trb.org/prp/16-1487.pdf
• Application Guidebook– Still in publication
Acknowledgements
Contact InformationEdward SmaglikAssociate ProfessorCivil and Environmental EngineeringNorthern Arizona UniversityEmail: [email protected]: 928-523-1431Sirisha KothuriResearch AssociateCivil and Environmental EngineeringPortland State UniversityEmail: [email protected]: 503-725-4208